This randomized, controlled effectiveness trial will test the effectiveness of a short-term, collaborative care model of cognitive- behavioral therapy for depression (CC-CBT) as an adjunct to anti- depressant medication treatment of adolescent major depression in primary care (principally pediatricians). The CC-CBT model, which is based on the co-management of a patient's treatment by a mental health specialist. and a primary care physician, has been developed and positively evaluated by co-investigators Von Korff, Lin, Ludman and Bush in a sample of depressed adults. In the first quarter of the initial study year we will pilot and revise the CC-CBT therapy manual for use with adolescents and families, prior to the randomized, controlled trial (RCT) in study years 01 to 03, with final follow-up completed halfway through Year 04. The RCT will be conducted in four major primary care clinics in the Northwest Division of the Kaiser Permanente (KP) health maintenance organization (HMO), with an overall membership of approximately 420,000, of whom approximately 47,590 (11.3%) are adolescents aged 12 to 18 (inclusive). We will enroll 150 consenting adolescents in this age range receiving incident (first time) pharmacotherapy for major depression from a KP primary care provider. Subjects will be randomly assigned to either (a) usual care depression pharmacotherapy as typically delivered by HMO providers; or (b) usual care pharmacotherapy plus six to eight individual sessions of CC-CBT delivered in a consult fashion by a trained mental health therapist (not the HMO pharmacotherapy provider). CC-CBT sessions are scheduled to occur in association with regular medication check visits, whenever possible. We anticipate that CC-CBT subjects will exhibit higher initial depression recovery rates, greater medication regime adherence, greater satisfaction with services, better social and academic functioning, lower subsequent depression relapse/remission. We predict that while initial HMO health care utilization and costs will be higher for the CC-CBT intervention will be more cost effective (e.g., units of improvement per unit of cost) compared to usual care.